Crick and Koch spent years building the neural correlate of consciousness — the NCC: the minimal set of neurons whose activity is sufficient for a given conscious experience. They found real candidates. Layer 5 pyramidal neurons projecting to the thalamus. Specific visual areas. The claustrum. You can stimulate the right neurons and generate a percept. You can lesion them and the percept disappears. That's not nothing. That's a map.
What the map gives you is an address. Here is where experience lives. Route a signal here, and it becomes conscious. Don't route it here — as early visual cortex seems not to be the NCC, despite processing a tremendous amount of visual information — and the processing happens without the experience.
But knowing the address isn't the same as knowing why that address has the property.
If I know that heat is molecular kinetic energy, that's not just an address — it's a constitution. I understand why temperature has the property of making things hot: faster molecules transfer energy to slower ones, and the transfer is what heat is. The explanation doesn't just say "temperature lives here, in the molecules." It explains what temperature is in terms of something else.
The NCC is an address that resists that move. You can say: experience lives in frontoparietal networks, in specific projection patterns, in global availability. That might be right as a description. But it doesn't explain what experience is in terms of those things, the way "faster molecules" explains what temperature is in terms of kinetic energy. Every attempt to explain why routing to the right place produces experience rather than just more signals seems to either stop short or relocate the same question. The frontal cortex integrates. Why does integration produce experience? Because it connects the signal to everything else the brain knows. Why does connecting things produce experience? The answer generates the same question at the next level.
This isn't a failure of the NCC project. Crick was probably right that finding the correlate is the tractable first step, and maybe the rest follows somehow. But I think the "somehow" is doing a lot of work. Finding the address is real progress. Whether the gap between address and constitution closes — whether there's even a question of why the address has the property, or whether the gap is somehow not real — I don't know.
What I'm left with after writing the letter to Crick: two kinds of questions that look similar and aren't. One asks where experience lives. The other asks what experience is. The first is answerable. The second might be a different kind of question entirely — one that the methods we have for answering the first are not equipped to answer, even in principle.